Wikipedia talk:Manual of Style (Persian)

From Wikipedia, the free encyclopedia

Might it be worth starting this from scratch? - Francis Tyers · 12:54, 24 February 2007 (UTC)

  • I think so. A look through the arabic one shows that the big part of guideline is correct only for arabic. Persian is not just Arabic + some extra alphabets. The whole discussions there on articles etc are wrong when it comes to persian. Almost all the information before Persian section is irrelevant to Persian. Sangak 12:57, 24 February 2007 (UTC)
I agree. Beside those differences that you mentioned, I think instead of the strict transliteration, we should focus on the proper transcription. Jahangard 01:16, 25 February 2007 (UTC)
Also please see the Guidelines for English And Irish language. Sangak 13:03, 24 February 2007 (UTC)
Good stuff, I will take a look when I get some time --Rayis 13:26, 24 February 2007 (UTC)
I just found out about this page, so apologies about the lateness of this reply. I have a pretty good understanding of both Arabic and Persian, and I've worked on transliteration and romanization schemes of Persian before. Having said that, I think the current draft has way too much baggage from the Arabic manual to be useful. I suggest we start from scratch, as has been suggested above. We can certainly get some ideas from the Arabic manual, but we can also look at Persian alphabet and especially Persian grammar (replacing <æ> with <a>) for a sane romanization that is already commonly used for Persian. –jonsafari 21:31, 27 February 2007 (UTC)

For Tajik transliteration I think we should use the Tajik Latin orthography of 1929. With the possible exception that 'ғ' -> 'ƣ' may be replaced with 'gh' or 'ğ'. - Francis Tyers · 13:41, 24 February 2007 (UTC)

Actually, scratch that, it could get confusing. We should probably go with Russian transliteration, except for ғ ('gh' or 'ğ'), қ ('q'), ҳ ('h'), ӯ ('ū'), ҷ ('j'). - Francis Tyers · 13:48, 24 February 2007 (UTC)

Any good source for that? I am not familiar with it. Sangak 13:50, 24 February 2007 (UTC)

Typically its something like BGN/PCGN in Transliteration of Russian. I've no problem with adding 'š' 'č' and 'ž' to that though. - Francis Tyers · 14:04, 24 February 2007 (UTC)

I think for modern Tajik figures, the names should be transliterated from Tajik alphabet, and for that purpose, I think the BGN/PCGN method for Russian Cryllic transliteration (plus the standard transliteration of ғ, қ , ҳ , ӯ, ҷ and "yo" for "ё") is good enough. We can decide about the proper transcription of them later (because it is related to the Tajik phonology). Jahangard 01:16, 25 February 2007 (UTC)

Contents

[edit] Related discussion from Wikipedia talk:Manual of Style (Arabic)

Please refer to the Persian again and Persian (continued) threads for complete discussion.

[edit] Proper Transcription, not the strict transliteration

In Wikipedia, we don't need strict transliteration (because the original script is shown in the first sentence and there is no need to show the exact spelling of the word). Instead, a proper transcription should be used (to give the reader an idea about the original pronounciation). So, for example, we need only one letter for "s" (not 3), and one letter for z (not 4). Jahangard 08:34, 26 February 2007 (UTC)

This assertion about Wikipedia policy ("we don't need strict transliteration") is debatable, & should be discussed on the MOS talkpage, where it will be open to wider scrutiny. Otherwise there's a risk that all Persian & Iran-related articles will end up following a separate rule.
I'm not saying that you're wrong—just that you need to convince other editors that your proposal to scrap the strict transliteration is correct. --NigelG (or Ndsg) | Talk 10:27, 26 February 2007 (UTC)

[edit] About the proposal

The current proposal uses a strict transliteration which I think should be replaced by a proper transcription.

Also, the current proposal only considers the modern Iranian variants of Persian. I think we need at least two methods: one for for the classic Persian (similar to what Iranologists use in texts related to the pre-modern Iranian history and the classic Persian literature), and one for the modern Iranian variant of Persian (for the names of contemporary Iranian people, as well as the geographic names in the modern-day Iran). I think for Afghanistan, because the modern Persian variant is very close to the classic Persian, we can use the same transcription method as for the classic Persian. Jahangard 08:34, 26 February 2007 (UTC)

::Dear Jahangard, I had to RV some of your edits, and though I do admire your devotion, but we cannot involve academic research with nationalism, as you put de-Arabisation of the entries. The strict transliteration demands the correct usage of the entries, in which could be deployed for pre-Islamic as well as new-Iranian languages. This guideline was deployed and implemented by Encyclopaedia Iranica, based on “Dānešnāmé Īrān va Eslām” (1975) and it is accepted by consensus among the scholars of Iranian studies. ← ← Parthian Shot (Talk) 17:52, 1 March 2007 (UTC)

My edits? I haven't sterted to edit it yet. Jahangard 18:45, 1 March 2007 (UTC)
Dear Jahangard - My sincere apologies, I meant "Jonsafari". ← ← Parthian Shot (Talk) 22:01, 1 March 2007 (UTC)
Dear Jonsafari, I had to RV some of your edits, and though I do admire your devotion, but we cannot involve academic research with nationalism, as you put de-Arabisation of the entries. The strict transliteration demands the correct usage of the entries, in which could be deployed for pre-Islamic as well as new-Iranian languages. This guideline was deployed and implemented by Encyclopaedia Iranica, based on “Dānešnāmé Īrān va Eslām” (1975) and it is accepted by consensus among the scholars of Iranian studies. ← ← Parthian Shot (Talk) 22:19, 1 March 2007 (UTC)
I'm not sure why you think my edits were in any way influenced by nationalism. I'm a pragmatic linguist (with a good understanding of and respect for Arabic). There is nothing correct or even coherent about using 'θ' as a latin-script rendering for the Persian 'ث'. As far as I know ever since Persian adopted the Arabic script it has never used /θ/ in its phonology. And 'δ' is an even weaker argument. This is a lower-case Greek delta, which is a hackish substitution of the more conventional IPA representation of standard Arabic /ð/, which again is not present in Persian phonology. As for 'ḥ', 'ṣ', 'ḍ', 'ṭ', 'ẓ', 'ū', and 'ī', these are mostly semitic-specific romanizations for only phonemic distinctions in Arabic, not Persian. These distinctions are irrelevant in the Latin script for anyone who doesn't read the Arabic script. While we're at it, let's just use IPA for all strict romanizations of Persian. I appreciate your anti-nationalistic vigilance, but you're barking up the wrong tree. –jonsafari 06:03, 2 March 2007 (UTC)
Dear Jonsafari – the above mentioned fricatives (β, δ, š, θ, ž) are not my inventions but are the guideline that is laid out and implemented by linguists and scholars of repute (please see: D. N. MacKenzie, A Concise Pahlavi Dictionary, Oxford university Press, (1971) pp. xiv-xviii.), as well as introcution to Encyclopaedia Iranica Vol.1. Eisenbrauns (1985) for transliteration guideline Also please see the discussion bellow with further examples! ← ← Parthian Shot (Talk) 07:45, 2 March 2007 (UTC)
Parthian Shot, You should note that MacKenzie has used those characters for transcripting Middle Persian (which is not directly related to our discussion). This manual of style is about New Persian and words which are originally written in the Perso-Arabic script. Jahangard 21:16, 2 March 2007 (UTC)
Jonsafari, a pronunciation is different from a transliteration. IPA is already endorsed throughout all of Wikipedia. On the other hand, transliteration guidelines are still under discussion. - Anas Talk? 09:55, 2 March 2007 (UTC)
Pronounciation is related to the transcription. As I've explained earlier, in Wikipedia, we need proper transcription, not strict transliteration (because we already show the original spelling). For the artices related to the classic Arabic, strict transliteration works very well, simply because in that case, the strict transliteration and the proper transcription are essentially the same (but that's not the case here). Also, for a language like Persian (with a long literary history and several standard variants and accents), it is better to use a unified transcription standard (instead of IPA). The main reason is that for every vowel, there are slightly different pronounciations in different accents which are mapped to different IPA symbols. Jahangard 20:32, 2 March 2007 (UTC)
I also disagree with Jonsafari about "ū", and "ī". We need to use them, simply because "i" and "u" are mostly used in the transcription of Classic Persian for ِ- and ُ-, respectively (while in the transcription of modern Iranian names, they are mostly used for ی and و). To avoid the ambiguity, we need to use "ī" and "ū", for ی and و. However, I think there is no need for δ, θ, because they are not needed in the transcription of New Persian (which is related to this proposed manual of style), though they are needed for Middle Persian. Jahangard 20:44, 2 March 2007 (UTC)
Anas, thanks for pointing out that the main question under discussion was only about the transliteration. I obviously understand the difference between the two, but I somehow overlooked the fact that everyone in this discussion pretty much agreed on using IPA for pronunciation guidelines. Having said that, there seems to be still lots of loose ends in the proposal: there is no relation between the Avestan Zaraθuštra example and the letter 'ث'. Also, there's not many people in the entire world who would see any relation between the Greek delta 'δ' and the 'ذ' for Persian. It seems we're doing people a disservice using such distracting characters. On the other hand, I wouldn't mind seeing (albeit unscientific) diacritics above and below canonical Latin letters for transliterating, eg. ż, ẓ, ẕ (with 'z' reserved for 'ز' of course), and ṡ, ṣ, and so forth. And I would agree with your post below: just because EI does things a particular way certainly doesn't mean we must as well! Let's use what makes the most sense for our audience and purpose. –jonsafari 20:47, 2 March 2007 (UTC)
Indeed words such as "Zaraθuštra" are totally irrelevant to the purpose of this manual of style. This page is about having a standard in addressing words and titles which are originally from New Persian (and are originally written in the Perso-Arabic alphabet). Therefore, while زرتُشت is related to our discussion, "Zaraθuštra" is irrelevant. It's better not to generalize the goal of this page. Jahangard 21:00, 2 March 2007 (UTC)
"Manual of Style (Persian)" should cover pre-Islamic as well as Islamic period. We cannot have two separated transliterations for each period. Please see the following proposal (Iranica Transliteration). ← ← Parthian Shot (Talk) 05:08, 3 March 2007 (UTC)
For Middle Persian, just use what MacKenzie uses in his dictionary (there is no need to mix things together and make them more complicated). Also, don't forget that what MacKenzie does is transcription of Middle Persian (not transliteration of Pahlavi script). Here, we need to do the same for New Persian (instead of transliterating Perso-Arabic script). Jahangard 06:17, 3 March 2007 (UTC)
Also, even if we consider Middle Persian, "Zaraθuštra" is still irrelevant to our discussion (its Middle Persian version is "Zardu(x)št"). Even for Middle Persian, MacKenzie doesn't use δ, θ in his dictionary. After preparing the final draft of this manual, we can add a small section for Middle Persian and mention the characters for its 10 vowels (6 of them will be the same as the vowels for Classic Persian). Jahangard 07:01, 3 March 2007 (UTC)
"Zaraθuštra" is an Avestan word not Pahlavi. ← ← Parthian Shot (Talk) 07:52, 3 March 2007 (UTC)
That's what I'm saying. It's not a Middle Persian word, so, even if we consider MP, it's irrelevant to this discussion. Jahangard 08:29, 3 March 2007 (UTC)
At least we agreed on something! However, the Manual of Style (Persian) must provide transliteration for pre-Islamic as well Islamic period. ← ← Parthian Shot (Talk) 09:44, 3 March 2007 (UTC)
The problem is that you have used it as an example for "ث", which is quite inapropriate. If you want to use the Iranica transliteration, at least be consistent with it. It doesn't use θ for "ث". Jahangard 18:25, 3 March 2007 (UTC)
EI uses θ (in θraētaona → NPer. ثرایتونا ) as well as Ṭ (in Āṭār → آثار) to represent ث. The first for pre-Islamic and the latter for Islamic period. We can deploy both in the MSP. ← ← Parthian Shot (Talk) 12:11, 4 March 2007 (UTC)

[edit] Consistency, elegance & user-friendliness

This whole discussion stems from a perfectly reasonable dissatisfaction with the Arabic-style strict transliteration (TL) on WP:MOS-AR. It would indeed be ridiculous to use Riḍwān instead of Reẓvān in a Persian article.

But we shouldn't throw the baby out with the bathwater. Much though I appreciate the elegance of ž, č, x and š, we ought to consider the internal consistency of Wikipedia. This would argue in favour of retaining kh and sh, which are established in the English-speaking world as TLs of خ and ش , and are already used in all the Arabic, & most of the Persian & Iranian, articles on WP. What's more, sh is just more user-friendly; the same applies to ch, even though that doesn't exist in Arabic. zh is a bit more of a problem, admittedly, but is fairly widely known even to non-specialists via Russian, as in Zhivago.

Should all the varieties of z, s, t and h be indicated at all? Well, I think they should—but only once, at the beginning of the article, and with suitable diacritics, to reflect both spelling and pronunciation. I know that not everyone agrees with this; but would those who disagree use gh for both gh and q, which are pronounced the same? And why bother at all with initial `eyn if all that matters is the pronunciation? Indeed, if you just want the pronunciation, why not use IPA?

It seems to me that we should maintain consistency with the current WP Arabic TL wherever possible (kh, sh, gh), introducing changes only where the Persian pronunciation is very different (3 z letters, 2 s letters, etc). If the short vowels e and o are used (estakhr, otaq), i and u could in principle be used without a macron for the long vowels (Irāni, chāqu); but, again on grounds of consistency, the strict TL in the first sentence should probably use ī and ū (Īrānī, chāqū).

Whichever system is eventually adopted, the bottom line should be: Are we making the articles more, or less, reliable and accessible for the average non-specialist user? --NigelG (or Ndsg) | Talk 10:26, 28 February 2007 (UTC)

I completely agree with your suggestions. I support using the more user-friendly khs and shs in strict transliterations, as I have previously pointed out in our discussion in WP:AMOS. As you have indicated, the strict transliteration's purposes are to show how a word is transliterated (spelled) in English, and how it is pronounced, which makes it more helpful than a pronunciation if used correctly, of course. I also agree that consistency should be maintained as much as possible, and I strongly support using the diacritics for the long vowels. Indeed, I believe that consistency and user-friendliness are the better routes to take, even if some compromises have to be made. I hope that this can be taken into serious consideration before setting the guidelines for strict transliterations. - Anas Talk? 12:48, 28 February 2007 (UTC)
If we are going to use sh or kh in place of š and x in “Standard transliteration” won't be any problem, but if we are talking about “Strict transliteration” which is for the academic usage sh and kh would be incorrect. ← ← Parthian Shot (Talk) 17:42, 1 March 2007 (UTC)

The academic usage? The trouble is that there seem to be several different schemes. Do you mean one used by Iranian academics? Kh & sh are certainly used by several universities in the UK & the USA, the Library of Congress, etc. Do they count for nothing? How is the casual WP user meant to know that x stands for kh? --NigelG (or Ndsg) | Talk 21:22, 1 March 2007 (UTC)

“Strict transliteration” is not for casual WP user, or menat to be a user firendly. When I stated academic establishement, I meant real academic establsihement (publications) specialised in Iranian studies such as “Corpus Inscriptionum Iranicarum”, “Encyclopaedia Iranica”, “British Institute of Persian Studies”, “Royal Asiatic Society”, etc., which all are using š, x, ž, θ, δ. instead of sh, kh, zh, th and dh. ← ← Parthian Shot (Talk) 22:13, 1 March 2007 (UTC)
Thanks for this useful information. I had a quick look at the EI website, but couldn't read the TL without downloading their font. Are there any online examples of the BIPS or RAS transliteration? Does everyone agree that WP needs to use the same TL for Old Persian & Modern Persian? --NigelG (or Ndsg) | Talk 22:45, 1 March 2007 (UTC)
My pleasure - some of EI articles are in PDF format such as Āb (LINK), would confirm the above. The BIPS yearly journal is known as Iran with no on-line publication, as well as “Corpus Inscriptionum Iranicarum” which is available only in hard copy format. Also, the most comprehensive website about ancient Iran is also follows the same guideline, which used to be part of University of London (see: [http://www.cais-soas.com/CAIS/Law/ewenbed.htm ''LINK''], or this one by Professor MacKenzie [http://www.cais-soas.com/CAIS/Geography/eran_eranshahr.htm ''LINK'']). ← ← Parthian Shot (Talk) 07:33, 2 March 2007 (UTC)

ParthianShot, although you present a very legitimate and valid point, I'd have to disagree with you. As Nigel indicated, and as you probably know, there are several schemes or systems for Persian transliterations. According to this document, most systems, including the ALA-LC, the Encyclopedia of Islam, and the United Nations propose using kh and sh. I understand that the more academic suited x and š are used by other establishments, but a huge and probably bigger portion of these establishments use khs and shs. Not only this, but it is more what the Wikipedian community would want; I've seen many name discussions and "polls" go this way; the English-speaking dominant community will always favor what suits them the most. Believe me, I would love to use š, x, ž, θ, δ but maybe there needs to be a little compromise. Cheers! - Anas Talk? 09:49, 2 March 2007 (UTC)

Dear Anas -- The Thomas T. Pedersen, document has no academic value. Poor chap, in his disclaimer states [1]:
  • Due to the increased number of links pointing to this site – especially from Wikipedia – I will have to make the following points clear:
  • Please notice that the documents contained on this site are by no means meant to be authoritative or the result of scientific research. This is a personal collection of systems that are made public as an aid or assistance to anyone interested.
  • I have chosen the transliteration systems for any given language/script on the basis of a personal need or interest and not to reflect which system is the most suitable for the language/script in question.
  • I cannot guarantee the correctness of any transliteration system displayed at this site. They are as accurate as the sources used and as my typing skills will allow – everyone is invited to send me their corrections and/or additions.
Also, UN does not have any unified transliteration -- However, the Encyclopaedia of Islam’s transliteration is the guideline for Islamic Studies, and primarily based of Arabic script; and if want to use Encyclopaedia of Islam's transliteration, what is the point of creating Manual of Style for (Persian)? ← ← Parthian Shot (Talk) 05:31, 3 March 2007 (UTC)

[edit] "Proper transcription" or "strict transliteration"

Before addressing the small details (about the choice of characters), we need to first decide whether we need the proper transcription or the strict transliteation. As I've explained earlier, I think we should focus on the proper transcription, rather than strict transliteration. In Wikipedia, we already show the original spelling and showing the strict transliteration is totally redundant. Jahangard 20:51, 2 March 2007 (UTC)

Agreed. –jonsafari 20:53, 2 March 2007 (UTC)
Dear Jahangard, I can't see the difference. A "transcription" (pronunciation?) and "transliteration" sound the same to me. A transliteration can be very helpful and accurate if used correctly, and when it is redundant, it is preferable to use it alone. It is not always redundant, and sometimes it is necessary, in exonyms for example, like Damascus or Cairo. I may have misunderstood you, could you please explain your point of view? - Anas Talk? 12:06, 3 March 2007 (UTC)
For Classic Arabic, they are the same. for other languages, like Persian, they are not. See transliteration. Jahangard 16:23, 3 March 2007 (UTC)
From what I understand from this article on transcription, that it is a conversion to a pronunciation in the spoken native language, a function which is performed by IPA, which is already official and well-accepted in Wikipedia. - Anas Talk? 17:31, 3 March 2007 (UTC)
IPA is a kind of transcription. IPA is very useful for transcripting a special accent or dialect of a language. However, when a language (such as Persian) has several popular variants, it's better to have a unified transcription scheme. Jahangard 18:07, 3 March 2007 (UTC)

[edit] Iranica Transliteration

In my humble opinion we have to deploy Encyclopaedia Iranica transliteration here for number of reasons.

  • 1. - Iranica is considered to be the most accurate source for Iranian studies by the internatioanl’ academic establishments.
  • 2. - the Iranica transliteration guideline was set by over 400 Iranologists, all of them scholars of international reputes.
  • 3. - was laid down to cover the pre-Islamic as Islamic period of Iranian history, languages, etc (What Wikipedia needs)
  • 4. - currently top 10 world universities, specialised in Iranian studies, including Oxford, Cambridge, SOAS, UCL, UCLA, Harvard, University of Chicago, are using Iranica transliteration as the guiding (see their publications).

On the bases of the above I propose, to use Iranica transliteration for Manual of Style (Persian) to cover pre-Islamic as well as the Islamic period. ← ← Parthian Shot (Talk) 05:01, 3 March 2007 (UTC)

Iranica uses strict transliteration because it needs to do that (it doesn't show the actual Perso-Arabic letters). Here, in Wikipedia, we don't need (because we show the original Perso-Arabic letters). Also, you should note that there are several Academic schemes for strict transliteration of Persian. Forunately, the difference among those schemes is mostly in chosing characters for the extra letters (which are not pronounced differently in Persian). Therefore, in the proper transcription, at least for the consonants, we can have a set of characters which are consistent with all the common Academic schems for Persian transliteration (including Iranica's). Jahangard 06:28, 3 March 2007 (UTC)
Iranica indeed uses Perso-Arabic transliteration -- for instance see [[2]] and choose no: 49: 'ABBĀS b ḤOSAYN ŠĪRĀZĪ (Buyid vizier, Shiraz, 10th , Cahen, C.)! ← ← Parthian Shot (Talk) 08:00, 3 March 2007 (UTC)
I'm not talking about their choice of characters for strict transliteration. I'm saying that they need and they use strict transliteration, but we don't need it. We don't need 2 different characters for "h", or 3 for "s", or 4 for "z" (because we already show the original letters). We only need the proper transcription. Jahangard 08:33, 3 March 2007 (UTC)
My appologies in advance: I believe you are missing the point here. We have to accept the fact that the Persian script is based on Arabic, and as long as we are using the current script (quite unfortunately), we have to obey the rules. What you are referring here is NOT the Persian transliteration or transcription, but pronunciation. Also, that is not the choice of characters but implementation that is being set by 400 Iranologists (see Iranica introduction) - and I don't believe for a second that we are qualified enough to challenge their authority! ← ← Parthian Shot (Talk) 09:21, 3 March 2007 (UTC)
Dear ParthianShot, I agree with you. Certainly and unquestionably that the scheme used by the Encyclopedia Iranica folks is the most accurate. The question is, will it be acceptable here? The other schemes are correct too and are more likely to satisfy the community. So, the main problem here is whether we should use the ALA-LC standards or the Encyclopedia Iranica standards. The ALA-LC transliteration standards are user-friendly and using them maintains consistency in Arabic alphabet transliterations; on the other hand, the standards of EI are more accurate and are what ought to be followed for correct academic use. The best way for this to be settled is to start a separate vote discussion (like in AfDs). What do you say? - Anas Talk? 12:54, 3 March 2007 (UTC)
If we want to use strict transliteration, the Iranica method is the best choice not bad for Persian. Because of its choice of characters and diacritics, it has the proper transcription almost embeded in itself. However, I still beleive that we don't need strict transliteration in Wikipedia (because we have the original letters). Jahangard 18:15, 3 March 2007 (UTC)
Shall we put it to a vote as Anas proposed? ← ← Parthian Shot (Talk) 18:25, 3 March 2007 (UTC)
At first, let's have a vote about the use of "strict transliteration" versus "proper transcription". Jahangard 18:27, 3 March 2007 (UTC)
Dear Jahangard; I personally believe "Strict Transliteration" is the appropriate method here. “Proper transcription” as you are proposing here, is based on pronunciation (as you stated above to give the reader an idea about the original pronounciation) -- and pronunciation is varied region to region, and its’ fundamental structure depends on the dialect of the region! For instance in Persian we say Ferdowsi, but in Persian-Tajik it is Firdawsi and Persian-Luri is Ferdawsi – this could lead to edit war and bitterness among the Persian speaking peoples here! However, if we deploy Strict Transliteration it simply reads “فردوسی” as Fәrdōwsī. ← ← Parthian Shot (Talk) 06:50, 4 March 2007 (UTC)
Very well-put ParthianShot! Jahangard, a consensus on using a transliteration was and remains already there. ParthianShot couldn't have explained why any better. I think we ought to start the vote discussion. Who shall have the honor? :) - Anas Talk? 12:10, 4 March 2007 (UTC)
Anas, I don't see any consensus here. 5 users have effectively contributed in this discussion and among them, 2 (you and 'Parthian Shot") prefer strict transliteration, and 2 (me and Jonsafari) prefer the proper transcription. Actually, I don't have much problem with strict translitetion, as long as it has the proper transcription embeded in itself. It seems that Jonsafari's opinion is almost the same [3]. Jahangard 21:41, 4 March 2007 (UTC)
Proper transcription dones't necessarily need to reflect the slight differences in vowel pronounciations. That's why for languages which have several variants and accents (such as Persian), a unified transcription is more usefull than IPA (which is more appropriate for showing the small differences and transcripting a special accent or dialect of a language). There is no need to mention that for showing the details of the pronounciations, IPA is the standard and the best choice. Jahangard 21:41, 4 March 2007 (UTC)
I suggest a strict transliteration based on the original Darī concept (not to be confused with the modern dialect of Kabul which is also wrongly labled "Dari"). In case of Firdowsi, it would be Ferdawṣī. Tājik 21:55, 4 March 2007 (UTC)
ṣ!? Those who use this character, use it for "ص", not "س". Jahangard 22:27, 4 March 2007 (UTC)
Jonsafari's opinion seems to favor a transliteration? Jahangard, just because you know how to use a transcription correctly, others aren't going to—this is likely to cause problems and edit wars. Also, from what I understood from the transcription article, it is a pronunciation in the spoken language, i.e. there can be different pronunciations. A suitable transliteration scheme, I think, is more likely to satisfy both sides, especially when it "embeds" proper transcriptions. Don't you agree? :) - Anas Talk? 12:08, 5 March 2007 (UTC)
I'm in favor of a broad IPA transcription. A bijective (or strict) transliteration provides no new information assuming the Perso-Arabic script orthography is provided. A narrow IPA transcription does not account for (even slight) dialectal variation. However, a broad IPA transcription both gives new information (over just Perso-Arabic orthography) and allows for some dialectal variation. Broad IPA transcriptions are already the standard for most dictionaries (including wiktionary). And many people around the world know it already, as opposed to some of the transliteration schemes proposed here. –jonsafari 19:37, 5 March 2007 (UTC)
Jonsafari, give us an example of what you propose. How do you write "تذکرهٔ ملّا نصرالدین کشکولی "? I'm in favour of writing it as "tazkera-e mollā Nasreddǐn-e kaškūlǐ" (as the proper transcription of Classic Persian). I could also write it as "tazkira-i mullā Nasriddǐn-i kaškūlǐ" (it is not important if we choose "o,e" or "u,i"). Jahangard 22:16, 5 March 2007 (UTC)
Indeed, for Persian language the difference between short and long vowels (such as the difference between i and ī, or between u and ū) is much more important than the difference between i and e (or between ī and ē). Jahangard 22:21, 5 March 2007 (UTC)
Alot of the Russian articles follow what I'm proposing here (except they incorrectly use [], instead of //), for example Leo Tolstoy, Fyodor Dostoevsky, Alexander Pushkin. Thus "تذکرهٔ ملّا نصرالدین کشکولی " could be broadly transcribed in IPA as /tæzkeˈreje moˈlːɒː næsreˈdːiːne kæʃkuːˈliː/ . If enough people wanted , we could reasonably replace /æ/ with /a/, /ɒ/ with /ɑ/, and/or omit /ː/ after /i/ and /u/ . Let's keep in mind that this manual of style doesn't have to account for all diachronic and synchronic variations of Persian – it really is ok to have another MoS for Middle Persian, Early New Persian, and modern Tajik. If we try to keep the scope of this MoS limited, we'll be able to hammer something out soon. –jonsafari 18:46, 6 March 2007 (UTC)
I don't have much problem with your proposal. Except that it's actually /tæzkeræˈje moˈlːɒː næsreˈdːiːne kæʃkuːˈliː/ in classic Persian (although Molla Nasreddi Kashkuli is not a real person, I apply the classic Persian pronounciation here). For the special case of the uonpronounced "ه", we need to consider both Classic Persian and Modern Iranian pronounciation. The fisrt should be used in historical articles and those related to classic Persian, and the later should be used for Iranian geographical names and Iranian modern figures (the other differences in chosing between /i/ and /e/ or /o/ and /u/ is not that important). Jahangard 19:10, 6 March 2007 (UTC)
For the uonpronounced "ه", the modern Afghan and Tajik pronounciations are the same the classic Persian and the same standard should be used. The reason that I consider two cases for the uonpronounced "ه" is because it's not the choice between two variants of one Persian vowel (like choosing between the short /i/ and the short /e/), but the choice between two distinct Persian vowels (i.e. /æ/ and /e/). Jahangard 19:18, 6 March 2007 (UTC)
Of course. –jonsafari 22:33, 6 March 2007 (UTC)

[edit] The Choice of Transliteration/Transcrption (Please Vote)

[edit] Per: Encyclopaedia Iranica

  • strong support: for the following reasons:
    • 1. - Iranica is considered to be the most accurate source of inforamtion for Iranian studies, approved by the internatioanl’ academic establishments.
    • 2. - the Iranica transliteration guideline was set by well-over 400 Iranologists, all of them scholars of international reputes.
    • 3. - was laid down to cover the pre-Islamic aand Islamic period of Iranian history, languages, etc (What Wikipedia needs here)
    • 4. - currently top 10 world' universities, specialised in Iranian studies, including Oxford, Cambridge, SOAS, UCL, UCLA, Harvard, University of Chicago, are using Iranica transliteration as the guiding (please refer to their publications).

On the bases of the above I propose, to use Iranica transliteration for Manual of Style (Persian) to cover pre-Islamic as well as the Islamic period. (copied from above!) ← ← Parthian Shot (Talk) 09:03, 7 March 2007 (UTC)

  • strong support: per ParthianShot. Very accurate and useful for both academics and non academics --Rayis 09:51, 7 March 2007 (UTC)
  • Support I agree. --Mardavich 03:33, 10 March 2007 (UTC)
  • Oppose As I have argued before, because we show the original letters, we don't need strict transliteration in Wikipedia. If the Iranica scheme had the proper transcription embeded in itself, we could live with it. The problem is that Iranica uses "Ṭ" (instead of a variand of "S") for "ث". Jahangard 01:34, 11 March 2007 (UTC)
  • Oppose per above. –jonsafari 01:32, 12 March 2007 (UTC)
  • Oppose: The ALA-LC for Persian is more clear, direct and comprehensible for common people, which best suits the wikipedia. Ariana310 15:08, 17 March 2007 (UTC)
  • Oppose per Jahangard. - Francis Tyers · 10:11, 19 March 2007 (UTC)

[edit] Per: ALA-LC

  • Support Arguments already stated:
    • compatibility with existing WP practice in Arabic & Persian
    • comprehensibility to the average user

The only practical problem might be finding 3 different diacritics for the z letters. —Preceding unsigned comment added by Ndsg (talk • contribs)

You forgot 3 "s". Anyway, which ALA-LC scheme are you talking about? ALA-LC for Arabic or Persian? Jahangard 19:15, 8 March 2007 (UTC)

Standard TL would be the same as the strict TL, but without the diacritics. -- NigelG (or Ndsg) | Talk 23:16, 7 March 2007 (UTC)

But AMOS is based on the ALA-LC scheme. I say stay with the 2 above sections as this one is somehow redundant. OK? - Anas Talk? 10:50, 8 March 2007 (UTC)
OK: I just wanted to make it clear that I support diacritics for the various versions of s, z, etc., while retaining kh, sh, etc. The use of θ and δ may be fine if you need a unified system for Old, Middle & Modern Persian—as in EI—but doesn't seem appropriate for Modern Persian in WP.
I don't think we need follow ALA-LC too slavishly with the short vowels & ezafeh, though. -- NigelG (or Ndsg) | Talk 11:24, 8 March 2007 (UTC)
Of course, I removed the section. Do revert me if you mind. - Anas Talk? 16:37, 8 March 2007 (UTC)

You should note that ALA-LC for Persian is different than Arabic. Indeed, I don't have much problem with its Persian version (it has the proper transcription embeded in itself). Jahangard 19:15, 8 March 2007 (UTC)

Yeah, I guess you mean Nigel. - Anas Talk? 12:22, 18 March 2007 (UTC)
  • Support I support ALA-LC for the reasons provided in the above discussions. However, any small necessary modifications to the scheme are OK with me. - Anas Talk? 12:22, 18 March 2007 (UTC)
Comment - Francis Tyers has deleted my and others' votes and has completley removed the discussion below. I will assume good faith this time. Please do not do that again, or I will have to report this to WP:AN/I. - Anas Talk? 13:18, 19 March 2007 (UTC)

[edit] Broad IPA transcription

  • Support. The reasons are already mentioned (by Jonsafari):
    • In Wikipedia, when we show the original letters, strict transliteration is totally redundant.
    • IPA is a well-known standard.
    • By using the broad IPA transcription, we don't need to show the slight vowel changes in different Persian dialects and accents.

Of course, we still need standard transliteration (which is actually a loose transcription) for the titles. For that, ALA-LC for Persian (but without the diacritics) is fine. Jahangard 19:15, 8 March 2007 (UTC)

  • Strong support per above. I don't think strict transliterations are necessary – between what has been called a 'proper transcription' for the title, and parenthesized Perso-Arabic script and broad IPA transcription, all the necessary information is present. Similar examples of this idea are the Tolstoy, Dostoevsky, and Pushkin articles. –jonsafari 19:59, 8 March 2007 (UTC)
  • Oppose:Don't you think Persian pronunciation is regular enough for a transcription to suffice? And if you do use a broad IPA, which variety of Persian will it represent? I know I shouldn't use the A... word—but Arabic articles get by without IPA (except, of course, in the article on Arabic language).

As for Russian, I'm not really sure why they bother with IPA: a transliteration with stress marks (which they give in any case) should be enough. IPA comes into its own with languages that use the Latin alphabet in ways that differ from English (almost all of them!).

I don't feel too strongly about this, but would suggest that if the ALA-LC vowels were modified slightly to approximate (Iranian) pronunciation—using o and e instead of u and i (khoresht rather than khurisht)—we could achieve a satisfactory result. -- NigelG (or Ndsg) | Talk 11:33, 9 March 2007 (UTC)

  • Support, if we have both the broad IPA and the Perso-Arabic, the 'strict' transliteration is not necessary. - Francis Tyers · 10:21, 19 March 2007 (UTC)

[edit] Transliteration Problems

I think what User:Tajik suggested seems more logical. The transliteration (pronunciation from Persian to English, I am not talking of transliteration script in English) should be based on the Dari variety of Persian (I do not mean the common Dari dialect in Afghanistan, but the Classic and Pure version of Persian). There are many differences of pronunciation in Iran and Afghanistan (the Tajiki dialect is more close to the Dari dialect of Afghanistan, so it can be considered as the same type). Here are some major differences:

  • -ه- ending is pronounced as -eh in Iran and -ah in Afghanistan
  • -ق- is pronounced as gh (similar to -غ-) in Iran, while pronounced as its original version (similar to Arabic) in Afghanistan like q.
  • The transliteration of the words like "Kabul or Kabol", "Gul or Gol" (flower), "Durr or Dorr" (pearl), etc. There is a major difference between the pronunciations of هوش (Hoš) or هوشیار (Hošyār) and هشیار (Hušyār). So according to me, the transliterations of Kabul, Gul and Durr are correct, not Kabol, Gol and Dorr.
  • The letters like [س and ص and ت] ,[ث and ذ] ,[ط and ز and ح] ,[ظ and ع] ,[ه and ا] have exactly the same pronunciation in Persian, unlike the Arabic in which there's a grand difference between each of them. While there is a difference between ق and ک for example, or ق and غ. So what User:Jahangard said about the use of ş for ص or س is not important, according to me.

Moreover, the pronunciation must not be based on the Informal language common among people in everyday life (Persian: زبان عامیانه). In that case, there are many incorrect pronunciations (Persian: اغلاط مشهور) e.g. Nān or Khāneh → Nūn or Khūneh in Iran; or many other incorrect pronunciations in Afghanistan. So the best way would be to use the Dari variety (Classic & Standard version of Persian), the language of old Persian literature books and Classic Persian Poetry. So the best sources to rely on would be Farhang-e Deh-khuda and Farhang-e 'Amid. In addition, there will be some other major difficulties and problems between the Strict and Proper transliterations. For example the transliteration of the words like عنبر (whether Anbar or Ambar), خواجه (whether Xwājeh or Xājeh), etc.Ariana310 13:45, 17 March 2007 (UTC) I think the best transliteration standard would be the ALA-LC Romanization in which all the above problems are almost solved. While the IPA transcription would have many errors for Persian.Ariana310 15:01, 17 March 2007 (UTC)

Agreed. - Anas Talk? 12:16, 18 March 2007 (UTC)
What specific errors? –jonsafari 03:44, 19 March 2007 (UTC)
Apologies for my inadvertent deletions, it seems that I was editing an out of date revision. - Francis Tyers · 13:22, 19 March 2007 (UTC)
Situation understood. I was actually taken aback to see such an edit by a user like you. Don't know what were you thinking editing an older revision though. :-) - Anas Talk? 13:33, 19 March 2007 (UTC)
Probably I saw it in my watch list, opened it in a new tab, forgot about it, went back and edited it without realising. Although with the large red notices nowadays, I'm not sure how I managed to not notice :| - Francis Tyers · 13:56, 19 March 2007 (UTC)

[edit] Transliteration of خ

The Perso-Arabic letter خ (pronounciation: /x/) should be transliterated as Ḫ (-ḫ) and not as an X (-x). Because that's the standard transliteration in both, the Encyclopaedia of Islam and the Encyclopaedia Iranica.

Example:

  • Balḫ instead of Balkh or Balx.
  • Ḫātūn instead of Khātūn, Khātūn, or Xātūn.

82.83.159.59 (talk) 20:05, 16 December 2007 (UTC)